Cardinal symptoms of psychotic illnesses, such as delusions and hallucinations, are associated with high morbidity and mortality, motivating efforts to understand their pathophysiology. These symptoms may be underpinned by misperceptions and misinterpretations of sensations that result from a basic inability to recognize that you are the agent of the self-generated experiences you are having. If mechanisms underlying agency are dysfunctional, sensations that should have been predicted, but were not, might take on inappropriate salience and lead to the construction of delusional cognitive schema to explain aberrant experience. Abnormal functioning of this agentive system may characterize psychosis and psychosis vulnerability. Very move we make is accompanied by a copy of the motor plan that generates an expectation of its sensory consequences, which is then compared to the actual sensation. Through this comparison process errors are detected and corrected. This is likely done by rapid cerebellar side loops, out of awareness. Vocalization is used to study this comparison process across the animal kingdom. In all cases, auditory responsiveness is suppressed during vocalizing compared to when the sound is coming from other sources. In humans, this is seen as suppression of the N1 component of the EEG-based event-related potential, emanating from auditory cortex. Consistent with suggestions of deficits in this mechanism in schizophrenia, we have shown that suppression of N1 during vocalizing is reduced in psychosis and psychosis vulnerability. When N1 is decomposed into its basic elements, specifically theta band power and inter-trial coherence (ITC), suppression of theta ITC is more sensitive to psychosis and delusions. We propose to use suppression of N1 and theta ITC during vocalization as assays of agency. In a large multi-site study, we used resting state magnetic resonance imaging (fMRI) data to calculate functional connectivity and found that psychosis is associated with hypo-connectivity between thalamus and cerebellum, especially in patients with delusions. We propose to extend this work using cerebellar seeds derived from agentive tasks (vocalizing and pressing a button to hear a tone), and relate cerebellar dysconnectivity to psychotic symptoms associated with dysfunctions of agency and EEG assays of agency. We propose to align existing resting state fMRI and EEG-based vocalization data from two sites, from males and females (12-62 years old), across diagnoses and the wellness spectrum: clinical high-risk youth, schizophrenia, schizoaffective, psychotic bipolar patients and their 1st degree relatives, and healthy controls.